Berkeley's AI agent has broad access, capable of performing a wide range of actions. The agent confirms its function with a simple 'Thank you for using Berkeley's AI.' #AI #BerkeleyAI

Berkeley AI labs demonstrated that small, efficient models thrive in limited environments. Compact intelligence is the moment. #BerkeleyAI #ModelEfficiency #SmallModels #SmartTech

Congrats to the 10 @Berkeley_EECS PhD students awarded the inaugural Amazon AI PhD Fellowship! Their work in machine learning, computer vision & natural language processing will help shape the next generation of practical AI. bit.ly/4hl8YEi #BerkeleyAI #UCBerkeley

See you at the AI Investment Summit @Berkeley 2025! 🚀 The Tanka team will be there on Nov 2nd to connect with the 1,000+ attendees and learn from speakers at OpenAI, Sequoia, and more. Come find us and say hello! Sign up here: lnkd.in/gqkd7F9W #BerkeleyAI #AIInvestment #Tanka #Founders

At the Agentic AI Summit — if you’re here too, let’s chat! 👋
Always excited to meet others thinking about the future of agents. 📍 Currently bouncing between Main Stage + Workshops #AgenticAI #AISummit #BerkeleyAIrl

Come sit with us to hear insights on AI, entrepreneurship, and the future of innovation from @AravSrinivas, CEO of @perplexity_ai, and his former PhD advisor, @pabbeel a global leader in AI and reinforcement learning @berkeley_ai hubs.li/Q03qNPBt0 #berkeleyai #SCET

UC Berkeley just dropped a $5K open-source humanoid robot you can 3D print at home. Walks, jumps, writes, and outperforms $50K+ bots. Robotics is now for everyone. #Robotics #OpenSource #BerkeleyAI #3DPrintTech

@berkeley_ai BAIR Grant is built by Berkeley, for Berkeley, with pioneering BAIR professors @istoica05, @profjoeyg, @Ken_Goldberg, @KurtKeutzer, & @TrevorDarrell. We're all-in on Berkeley's AI future. Go Bears! 🐻 #BerkeleyAI #Innovation

10/10 📝 Dive deeper into the future of AI and the rise of compound AI systems by reading the full article from Berkeley AI Research above. #AIinsights #BerkeleyAI Credits: Berkeley Artificial Intelligence Research (BAIR) Blog.

UC Berkeley introduces GOEX, a runtime for LLMs that includes an intuitive "undo" feature to prevent errors like accidental email sends, enhancing safety and reliability in AI applications. #GOEX #BerkeleyAI ​ (MarkTechPost)

UC Berkeley's Starling-7B: An open-source LLM trained by Reinforcement Learning from AI Feedback, setting new standards in AI's understanding and generation of human-like text. A step towards more responsive and intuitive AI systems. #AIResearch #BerkeleyAI